Unsolved
This post is more than 5 years old
11 Posts
0
5528
August 3rd, 2011 13:00
Issue with extraneous paths showing up in Powerpath
I have an Exchange cluster (Windows Server 2008) with 3 luns from my CX700 and 1 from my CX4-480. There should be 8 paths zoned between the server and the CX700. Looking in Powerpath on 2 of the luns going to the CX700 I was actually seeing 10 paths. The 8 expected that were optimal and 2 additional paths that were dead. These were pointing to ports SPA1 and SPA2. What I noticed was the target portion of the pathname was incorrect on the 2 dead paths. First we upgraded from Powerpath 5.2 to 5.5 hoping that might resolve the issue. It didn't. I concluded that since the 8 paths we have zoned were showing as optimal the other 2 must have been just glitches from some time past. I ran powermt check to remove them then powermt save. Dead paths went away. Looked like life was good.
Then I logged onto the server again about 3 ours later and founf I now had another dead path only this time it was pointing at SPB0.
Has anyone experienced an issue like this? I think functionally everything is working ok on the server, it just keeps getting these extraneous paths showing up.
RobertSnyder
11 Posts
0
August 4th, 2011 08:00
C:\Program Files\EMC>powermt display dev=all
Pseudo name=harddisk1
CLARiiON ID=APM00051301096 [EXCA-AK-FIS]
Logical device ID=600601603D741600CE935A711654DE11 [LUN 340]
state=alive; policy=CLAROpt; priority=0; queued-IOs=0;
Owner: default=SP B, current=SP B Array failover mode: 1
==============================================================================
--------------- Host --------------- - Stor - -- I/O Path -- -- Stats ---
### HW Path I/O Paths Interf. Mode State Q-IOs Errors
==============================================================================
4 port4\path0\tgt0\lun0 c4t0d0 SP A0 active alive 0 0
4 port4\path0\tgt2\lun0 c4t2d0 SP A2 active alive 0 0
4 port4\path0\tgt3\lun0 c4t3d0 SP B3 active alive 0 0
4 port4\path0\tgt5\lun0 c4t5d0 SP B1 active alive 0 0
5 port5\path0\tgt0\lun0 c5t0d0 SP A1 active alive 0 0
5 port5\path0\tgt2\lun0 c5t2d0 SP B2 active alive 0 0
5 port5\path0\tgt3\lun0 c5t3d0 SP A3 active alive 0 0
5 port5\path0\tgt5\lun0 c5t5d0 SP B0 active alive 0 0
Pseudo name=harddisk2
CLARiiON ID=APM00051301096 [EXCA-AK-FIS]
Logical device ID=600601603D741600CF935A711654DE11 [LUN 341]
state=alive; policy=CLAROpt; priority=0; queued-IOs=0;
Owner: default=SP B, current=SP B Array failover mode: 1
==============================================================================
--------------- Host --------------- - Stor - -- I/O Path -- -- Stats ---
### HW Path I/O Paths Interf. Mode State Q-IOs Errors
==============================================================================
4 port4\path0\tgt0\lun1 c4t0d1 SP A0 active alive 0 0
4 port4\path0\tgt2\lun1 c4t2d1 SP A2 active alive 0 0
4 port4\path0\tgt3\lun1 c4t3d1 SP B3 active alive 0 0
4 port4\path0\tgt5\lun1 c4t5d1 SP B1 active alive 0 0
5 port5\path0\tgt0\lun1 c5t0d1 SP A1 active alive 0 0
5 port5\path0\tgt1\lun1 c5t1d1 SP B0 active dead 0 0
5 port5\path0\tgt2\lun1 c5t2d1 SP B2 active alive 0 0
5 port5\path0\tgt3\lun1 c5t3d1 SP A3 active alive 0 0
5 port5\path0\tgt5\lun1 c5t5d1 SP B0 active alive 0 0
Pseudo name=harddisk3
CLARiiON ID=APM00051301096 [EXCA-AK-FIS]
Logical device ID=600601603D741600C0269E741D54DE11 [LUN 314]
state=alive; policy=CLAROpt; priority=0; queued-IOs=0;
Owner: default=SP A, current=SP A Array failover mode: 1
==============================================================================
--------------- Host --------------- - Stor - -- I/O Path -- -- Stats ---
### HW Path I/O Paths Interf. Mode State Q-IOs Errors
==============================================================================
4 port4\path0\tgt0\lun3 c4t0d3 SP A0 active alive 0 0
4 port4\path0\tgt2\lun3 c4t2d3 SP A2 active alive 0 0
4 port4\path0\tgt3\lun3 c4t3d3 SP B3 active alive 0 0
4 port4\path0\tgt5\lun3 c4t5d3 SP B1 active alive 0 0
5 port5\path0\tgt0\lun3 c5t0d3 SP A1 active alive 0 0
5 port5\path0\tgt1\lun3 c5t1d3 SP B0 active dead 0 0
5 port5\path0\tgt2\lun3 c5t2d3 SP B2 active alive 0 0
5 port5\path0\tgt3\lun3 c5t3d3 SP A3 active alive 0 0
5 port5\path0\tgt5\lun3 c5t5d3 SP B0 active alive 0 0
Pseudo name=harddisk4
CLARiiON ID=APM00092901686 [EXCA-AK-FIS]
Logical device ID=600601600F3022000C742C39F64BE011 [LUN 580]
state=alive; policy=CLAROpt; priority=0; queued-IOs=1;
Owner: default=SP B, current=SP B Array failover mode: 1
==============================================================================
--------------- Host --------------- - Stor - -- I/O Path -- -- Stats ---
### HW Path I/O Paths Interf. Mode State Q-IOs Errors
==============================================================================
4 port4\path0\tgt1\lun0 c4t1d0 SP B1 active alive 0 0
4 port4\path0\tgt4\lun0 c4t4d0 SP A0 active alive 0 0
5 port5\path0\tgt1\lun0 c5t1d0 SP A1 active alive 0 0
5 port5\path0\tgt4\lun0 c5t4d0 SP B0 active alive 1 0
c5t1d1 and c5t1d3 are the bogus paths. Those don't exist and I don't know why Powerpath thinks they are there. Previously I had c5t3d1, c4t1d1, c5t3d1, and c4t1d3 associated with A1 and A2 that were showing up. I ran the check command to remove them and it looked good for 2 or 3 hours then these paths showed up on B0.
Julien_LECORRE
92 Posts
0
August 4th, 2011 08:00
Can you run powermt display dev=all and provide the output ?
dynamox
11 Legend
•
20.4K Posts
•
87.4K Points
0
August 4th, 2011 09:00
does the target change everytime they come in and out ?
Julien_LECORRE
92 Posts
0
August 4th, 2011 09:00
Yep sound strange...
Have you check the PowerPath_V55.log file ? Nothing special inside ?
RobertSnyder
11 Posts
0
August 4th, 2011 10:00
Where do I find the log file? Do I have to do something to turn logging on? I looked under the program files\emc folders but I don't see any log files there.
RobertSnyder
11 Posts
0
August 4th, 2011 10:00
The target has been different on each occurrence. It is the target portion of the address that seems to be the issue.
dynamox
11 Legend
•
20.4K Posts
•
87.4K Points
0
August 4th, 2011 11:00
you don't see these issues with CX4 LUNs ?
RobertSnyder
11 Posts
0
August 4th, 2011 11:00
No the only issues have been on the luns coming from the cx700.
I am having the same issue on the other Exchange clusters in the organization too. Hoping to get it fixed on this node that is currently the passive node and that that fix will resolve the other issues too.
RobertSnyder
11 Posts
0
August 4th, 2011 12:00
Thanks I found it.
Looks like an install log. I scanned through it and don't see any errors. Says install complleted successfully at the end.
dynamox
11 Legend
•
20.4K Posts
•
87.4K Points
0
August 4th, 2011 12:00
have you looked in CX700 SP logs, anything stands out over there ..like trespasses or anything like that ?
dynamox
11 Legend
•
20.4K Posts
•
87.4K Points
0
August 4th, 2011 12:00
i found PowerPath_V55.log file in C:\Users\
dynamox
11 Legend
•
20.4K Posts
•
87.4K Points
0
August 4th, 2011 13:00
did you install the P01 patch ?
RobertSnyder
11 Posts
0
August 4th, 2011 13:00
Actually we haven't. We just upgraded to the straight 5.5 release. I wondered about the patch. I actually have a support case open with EMC regarding it too but am getting nowhere. Was hoping maybe someone on the boards here had experienced something similar. Guess part of the reason I hadn't applied the patch is I was thinking if there was anything in it that addressed the issue support would have pointed that out. Maybe I'll go see if I can find what was fixed in p01.
RobertSnyder
11 Posts
0
August 4th, 2011 13:00
Everything looks good on the storage side. I'm really wondering if it's a powerpath bug of some sort.
kelleg
6 Operator
•
4.5K Posts
0
August 4th, 2011 14:00
have you tried running "powermt check"
you might need to reboot the host to fix the paths
glen